School Zoning System for Student Admission using Constrained K-Means Algorithms

Abstract

The issuance of the Regulation Minister of Education and Culture number 51 of 2018 regulates new student admission by implementing a zoning system to achieve equal distribution of education quality in every school, especially at the high school level in Makassar city. This study aims to cluster the school zoning area based on the closest distance between the student's domicile and the school location. The dataset used is 22 school locations and 2248 student location data. In this paper, the method used is constrained k-means to cluster the prospective new students to each school. The constrained k-means method works based on the value of K as the closest cluster center to the value of N (cluster members) with a linear programming algorithm (LPA) approach so that each cluster has a balanced N member. The results of this study can overcome the unbalanced data distribution problem with an average cluster member value of 103 and the absence of empty clusters in each school/centroid. Thus, the system can be implemented in the new student admissions process as a reference in determining the optimal and accurate school zoning area based on the cluster center.